RF Ranging Methods and Performance Limits for Sensor Localization

RF Ranging Methods and Performance Limits for Sensor Localization

Steven Lanzisera (University of California, Berkeley, USA) and Kristofer S.J. Pister (University of California, Berkeley, USA)
DOI: 10.4018/978-1-60566-396-8.ch004


Localization or geolocation of wireless sensors usually requires accurate estimates of the distance between nodes in the network. RF ranging techniques can provide these estimates through a variety of methods some of which are well suited to wireless sensor networks. Noise and multipath channels fundamentally limit the accuracy of range estimation, and a number of other implementation related phenomena further impact accuracy. This chapter explores these effects and selected mitigation techniques in the context of low power wireless systems.
Chapter Preview

Application Requirements

The requirements of a localization system are dependent on the application. This section will discuss a few applications to determine requirements on accuracy, latency, useful range, and infrastructure complexity of a ranging system. The accuracy requirement is defined to be the maximum error between true and estimated position that is acceptable for some percent of all estimates. For example, if 80% of estimates must be accurate to within 2 m, then 20% of measurements can have larger error. It is important to understand that localization is probabilistic in that the environment among other factors randomly degrades the accuracy of a measurement. Latency is the time it takes from when a request for a location update is made to when the update is presented to the user for a single device in the network. The range requirement is roughly how large of a sphere must one make around any node to find at least 4 other nodes or infrastructure points in 3D and 3 infrastructure points in 2D. Infrastructure requirements impact the cost of a network, and this impact can be considered qualitatively.

Complete Chapter List

Search this Book: